IntelRealSense / librealsense

Intel® RealSense™ SDK
https://www.intelrealsense.com/
Apache License 2.0
7.6k stars 4.83k forks source link

How to detect objects using the depth from the intel realsense in c# #7295

Closed Xander-Vreeswijk closed 4 years ago

Xander-Vreeswijk commented 4 years ago
Required Info
Camera Model D455
Firmware Version (Open RealSense Viewer --> Click info)
Operating System & Version Win 10
Platform PC
SDK Version 2.<?>.<?>
Language C#
Segment {Robot/Smartphone/VR/AR/others }

Issue Description

I need to detect different kind of things in a crate like leaves. For that problem I need to detect the objects using the depth image form the intel realsense. I have seen different kinds of solutions for c++ but not one for c#. Can someone help me with this problem.

MartyG-RealSense commented 4 years ago

Hi @Xander-Vreeswijk During research of your question, I found that it is much easier to find relevant search results about c# object recognition when googling for object detection c# deep learning, as virtually nothing is found when googling for the term object recognition c#.

There is the potential to integrate recognition solutions that you find into a librealsense application by using the SDK's compatibility with platforms such as OpenCV.

Below is an example from search results gained with the recommended search term of a C# system for detection and identification with a live camera:

https://github.com/fralik/ObjectDetection-MLNet

Xander-Vreeswijk commented 4 years ago

Thank you for your response. I will look if it works.

MartyG-RealSense commented 4 years ago

Hi @Xander-Vreeswijk Do you require further assistance with this case, please? Thanks!

Xander-Vreeswijk commented 4 years ago

MartyG yes do you know an other way of detecting objects using intel realsense because the microsoft.ml nugget package doesn't work on my laptop. Do you perhaps have a way of using opencvsharp?

MartyG-RealSense commented 4 years ago

There is a set of OpenCVSharp samples at the link below. Sample number 16 is a program that describes itself as "train your own detector", though it is likely not based on depth estimation.

https://github.com/VahidN/OpenCVSharp-Samples/tree/master/OpenCVSharpSample16

A RealSense user also suggested the following:


You could try using a third-party library like OpenCV. There's an example / tutorial for object tracking here:

https://www.learnopencv.com/object-tracking-using-opencv-cpp-python/

You can use OpenCVSharp to get it working in dot net.

Xander-Vreeswijk commented 4 years ago

thank you for your respons MartyG. I will try it

MartyG-RealSense commented 4 years ago

Hi @Xander-Vreeswijk Do you require further assistance with this case, please? Thanks!

Xander-Vreeswijk commented 4 years ago

No it is clear thank you for your assistance